TECHNISCHE UNIVERNTET Laboratorium voor Scheopshychomodianki
Archlet
Mekehvog 2, 2828 CD Den TOL' OM- 786873 Noe Ma nian
PROCEEDINGS
NINTH
SHIP CONTROL SYSTEMS
SYMPOSIUM
10-14 SEPTEMBER, 1990
BETHESDA, MARYLAND, U.S.A.
VOLUME 2
A Portable Automatic Control System for Ocean Research Operation of a Ship with
a Controllable Pitch Propeller, a Rudder and a Bow Thruster by
HIROYUKI ODA
Akishima Laboratories (Mitsui Zosen) Inc. and
KIYOSHI MASUDA & KEIICHI KARASUNO Hokkaido University, Fisheries
ABSTRACT
The fisheries training ship " Oshoro Maru (1,400 G.T.) " is
equipped with a controllable pitch propeller, a rudder, and a
bow thruster. They are used in combination for both course
keeping and speed control of the ship while
conducting
its oceanresearch operation.
While
conducting
ocean research operations, manually operatingthe ship via the actuators mentioned above, is troublesome due
to strong external influences such as wind, waves and current.
Adding to the problem is the limited number of actuators and so
on. In order to solve these problems, a portable automatic
ship control system for ship operation has been designed using
computer simulations and full scale trial at sea.
The Portable Automatic Control System (PACS) consists of
pre-programmed micro computer modules for control, operation, and
I/O. This system makes it possible to directly select several
combinations
of actuators from
console panels located thebridge deck and in the research room.
Results obtained so far, indicate this control system is able
to reduce operator weariness more than the conventional manual
cruise operations. This paper describes the implemented system
with particular emphasis on the control system.
1.
Introduction
Fisheries
training
ship " Oshoro Maru " of Hokkaido Universityand also fisheries
training cruises in the North and South
Pacific Ocean. [1]
In the case of oceanographic research, CTD equipment and
sounding machine are dropped over the ship's side by use of the
very long cable (more than 3,000m). This is done so the crew
can operate her well at dead slow speeds using the actuators of a Cpp, a rudder, and a bow thruster to keep her course and speed
steady.
This prevents measurement equipment from
becomingdamage due to cable tanglement under the ship.
However,
manually operating these ship actuators is
very
troublesome against the external influences such as wind, waves and current. This situation is considerable mental and physical pressure to the operators. In order to solve these problems, we developed a portable automatic controller with particular emphases on keeping the ship's course and speed steady in dead slow speed or under stopping
conditions
against the externalinfluences. The main design consideration is to keep both the
ship and the instruments safe during ocean research operations.
An approach to achieve this control system and its designing
information
is described in Section 2. The computer simulationdesign and mathematical model are summarized in Section 3. The
control system design is reviewed in Section 4. The results of
full scale test and several problems occurred at full scale test
are given in Section 5. Finally, in Section 6 it is concluded
that the ship can be automatically controlled at
smooth andmedium sea
condition,
but in rough sea condition, it is
necessary to adjust the control low and bow thruster for the
endurance problems. It is also concluded that this automatic
control system is safer than skilled manual control
whilecourse keeping at dead slow speeds or under stopping
conditions.
2. Design Approach
2.1 Oceanographic Operation
"
Oshoro Maru
" isspecially
designed
for fisheries
training
and oceanographic research. The ship is designed with a stern trawler equipped with a Cpp, a rudder, and a bowthruster. The particulars of the ship are summarized in Table 1.
An outline of the ship's structure and the photograph are shown
in Fig.1 and Photo.l.
The maneuvering image during
oceanographic observation, which may suggest basic strategies of
the automatic control system, is shown in Fig. 2.
At thestopping conditions, the long cable deploys the observation
Length(o.a)
LengiNp.p)
Breadth(mId) Depth(m1d)2nd
Upp. Dk Full loaded draft Gross tonnage Japan/Int. Call sign
Main engine Hanshin
Propeller Main generator Dow thrustor Anti rolling tank Trial max. speed
Service speed(15% load)
Number of complement
-Photo. 1 View of "Oshoro Maru"
Table 1 The particulars of "Oshoro Meru"
To prevent the cable from tangling under the ship or into
the screw propeller during the observation, the ship was
controlled as described below:Get the wind on the port side of the ship about 10 deg. ahead.
Set the ship in
hovering position or with only a small
movement astern.
131 Tuna long line
I I ydraulic line hatalor 0.2t 272m/min. > 1 tel
tit Salmon gill net
llydraulic net hauler 0.3t x I77m/min.
Oceanographic winch Sounding machine
Hydraulic 0.6t x 120m/mm. 4 41 x 5.000m Set
121 C.T.D.winch
Hydraulic It x 72m/min. 6.4 x 4.000m x I set 131 Underwater T.V. winch
Electric 0.5t x 3 12m/min. 20 z 500m X I set
Ill Sampling winch
Hydraulic 2.5t a 120m/min. 12 is x 3.200m X 1 yet
Laboratory
Computer. Disk. Graphic plotter. Printer. Aquarium for culture fishes. Thermu.salinometer.
Fishing .equipment
Soner, Fish finder. Net recorder. Qamitative echo sounding system. Otter trawl dull, logging and analyzing system
Navigation equipment
Gyro compass. Radar x 2. Anti collision system. Hyorid navigation system (LoranA &C.Omega. N.N.S.S.Decca processor). Decca. Electro magnetic log. Ultrasonic anemometer
Fishing gear
Ill Trawl winch
hydraulic lOts 80m/min. 22 # 3000m I set
(21 Net carrier
Hydraulic 0.31 X 100m/min. S 2 sets 72.S5m 66.00m 12.60m 3.40m 5. itint 5.00m 1341301779i it/VA 6EL40 3.2009.s. x 240rprn x 1 set
4 -Waded C.1'.I'. x I set
4501i V A x 3 sets
Thrust5.5t. 390kW x I vet U duct type x 1 set
14.43knots 13.4knots
Officers 139. Crews 27p. Scientists 6p. Cadets 609.
In this manner, if the ship is off its desired performance,
then her course and speed would
be adjusted by a Cpp and
a
rudder. However, if this does not produce sufficient results
then the bow thruster can be employed in cooperation with a
Cpp and a rudder without chattering action.
Fig. 1 An outline of "Oshoro Meru"
2.2 Control Dynamics
The basic problem for the realization of a controller is how
to obtain practical information useful to the object system. One
of the method is the statistical approach which makes use of a
parametric model that is obtained by
the data analysis. Thedesign is based on the understanding of the system behavior which is obtained by careful statistical analysis of the actual data. By using an appropriate statistical model, this approach
provides a model that takes into account both characteristics of
the system and noise sources. The detail of this method is
discussed by the authors. [3][4]
Here we introduce the analysis of numerical data observed by
an actual experiment at sea. Fig.3 shows a portion of head
angle, rudder angle and pitch angle of thruster. Fig.4 shows
the decomposition of power of yaw angle, rudder angle, and pitch
angle of bow thruster in relation scale.
Fig.5 shows the
relative power contribution of system variables. These figures
show that the contribution from rudder
and bow thruster are
significant in the low frequency band.
These results show the effects of some definite feedback
relations occurring within the controlled variables. These are
significantly amplifying the frequency
components mentionedabove within the random sea. [5]
MIND.VAVE(11 CURRENT EQUIPMENTS CABLE E.--BOW THRUSTER am. <4---CPP RUDDER
11111 awn's.
Fig. 3 Observed data ( Actual Sea Test )
TAY
Fig. 4 Power spectrum
m cm em om om om
IA 1
MOM*
0.4M
ON.
Fig. 5 Relative power contribution
2.3 Design Concepts
In aid of the understanding for
the configuration of the
proposed Portable Automatic Control System
(PACS), a briefdescription of each concept is given below:
The PACS introduces a step forward control of the shipboard,
by combining the reliability
and flexibility of digital
computers with the simplicity of control procedures.
The manual control of actuators mentioned above, leads to
complex operation adjustment. The PACS may greatly decrease
this troublesome job.
The PACS is able to select the mode of the manual control or
the automatic control for each actuators, and in the automatic
control mode, to select combinations of actuators separately
depending on sea conditions.
The PACS is able to set in suitable position on board, for
example steering room in the bridge deck, wing deck or research
room. Me OM OW"' M CM OM OM OM IOU 11101 TO 100011 0 THROSTIR SM eM 4.0 MUM OM 10100111 OM OM 100010 O. IMO 'TAW
o
The organization of PACS as a whole, is represented by Fig.6::
The PACS can select the control mode of manual or automatic in a
bow thruster, a Cpp and
a rudder, respectively as shown in
Fig.7. The desired direction is controlled with a bow thruster',
and a rudder depending on the Cpp movement.
Also the desired' velocity is controlled with only a Cpp.
3. Portable Automatic Control System
3.1 Function and Operating
INPUT GYROCOMPASS SPEED LOG WIND SENSOR BOW THRUSTER Cpp RUDDER PROCESSOR UNIT OPERATION BOX OUT
Fig. 6 Organization of PACS
Fig. 7 Control mode
We show specific control modes below: * Full Automatic Control
In this case, a bow thruster,
a Cpp, and a rudder are
selected in auto mode. A desired velocity and heading
angle of the ship is set with the dial on the operator's panel. A pitch
angle of the bow thruster, Cpp, and rudder angle are calculated
by the main computers in the control console so as to keep her
velocity and heading angle. This is based on the signal
from the doppler log and gyro compass. The desired direction are
controlled with a bow thruster and a rudder dependent on a Cpp
movement, and the desired velocity
is controlled with onlya,
Cpp.
* Manual Control
In this case,
a bow thruster, a Cpp, and
a rudder are
selected by manual mode. The pitch angle of bow
thruster and
Cpp are selected in proportion to the inclination angle of the
joystick. The rudder angle is ordered by the indication angle of
the dial.
In this mode the
crew operates the ship with the
joystick lever and the rudder dial.
BOW THRUST( COD RUDDER 0 : POSSIBLE X : IMPOSSIBLE IBOW MUSTER SET UP HEAD x 0 ---x .--0 -.. 0 ... 0 ... 0 0 SPEED x x 0 0 x x 0 0 AUTO. CONTROL THRUSTER X x x X 0 0 0 0 Cpp x 0 0 X X 0 0 RUDDER X 0 X 0 X 0 X 0
* Semi-Automatic Control
In this case, a bow
thruster, a Cpp, and a
rudder are
respectively selected by manual or auto mode. For instance,
when a bow thruster is selected in auto mode, and besides a Cpp
and a rudder in manual mode, a desired direction may be set on the operator's panel. Then a pitch angle of bow thruster is
calculated by the computers so as to keep the
direction.
Furthermore, a Cpp and a
rudder may be controlled with a
joystick and dial freely.
The control sequences mentioned above are shown in Fig.8.
Fig. 8 Control sequences
3.2
System Configuration* Hardware
PACS consists of two components; one is a processor unit with
a 16-bit CPU, and the other is a portable control unit with a
joystick and dial. Photo 2 shows the appearance of the portable
operation box with its joystick, dial and several buttons and
display units.
A single on-board
computer(MAC6000) and theappearance of the processor
unit are, shown in
Photo.3.A
special feature is the button switches located on the right side
of the joystick which enables easy mode change.
Once the mode is changed, the ship's heading and
speed can be
controlled by selecting the combination mode.
This can set each actuators on this operator's panel in automatic or manual mode. An operator's panel and its faculties are illustrated in
Fig.9. The ship is already equipped with a manual
operation
apparatus on the operation
console in the bridge.PACS was
installed
by setting the CPU
control circuit
as shown in
Fig.10, and by using the original control apparatus as much as
possible.
HEAD SETTING SPEED SETTING
DIAL (RUDDER)
Fig. 9 Operation panel
Photo. 3 Processor unit
JOYSTICK(THRUSTER. a Cpp)
WIND INDICATOR
COHPUTER ROOH
I/O BOX PACS
MISR BRIDGE ISWITCH Cpp chit v I I OW THRUSTER RUDDER HYBRID NAVIGATION SYSTEM
Fig. 10 Ship control system
It was judged very important that all in/outgoing signals are
correctly insulated and isolated one from another. For this reason, every input/output signal coming from, or going into,
the PACS is processed according to the scheme of Fig.11. * Software
The implementation of the software design was made bearing in
mind the follows:
an ability to deal with large program size, quick software
analysis, design, writing and debugging, minimizing at faults
during real operation.
possibility to model the software easily
during sea
trials.
For these purposes the application development on the MAC6000
was done in a greatly simplified manner by using high-level
languages and functions. Development aids were prepared and run-time communication support was conducted.
The MAC6000 control program for the PACS is divided into four
parts, and constructed with the following functions: Disposition of extreme values
Filtering of input data
Modified multi variate PID control Wind compensation
Distribution of calculated power to several actuator Management of abnormal movement of several actuator Monitor and printing functions.
The force allocation sequences are as follows:
The total forces and
moment which are required from the
regulators, must be converted to controller, rudder, and Cpp commands. This is done in the force allocation algorithm with and without the use of a bow thruster. The latter mode is used
for the ship operation requiring less vibration
and noisereduction. In this mode the controller commands to a Cpp and a
rudder are computed to give force and moment balance. The bow
thruster is used for operating at low speed and for hovering.
In this force allocation mode, the controller commands to the
Cpp are not only determined for an axis force, but also for the
rudder turning moment.
If, however, it is not possible to
obtain a sufficient moment, the bow thruster is also controlled
together with the rudder. The commands to the
rudder are
ordered, if the Cpp is going to a positive pitch, even though the corresponding rudder command is always zero
if negative
pitch.Details of the program structure control is presented in Fig.12. AC 1000 PACT 11100 POOH 0111 PE0ATIO4 DM OPMMIIM .MMCA DIAL .MTCA MM 4111110 NAVIGATIOA SIXTH (railm) 0CORD !MR CM INIAiaANCA) NEADING.CUMALTIME VINO VELOCITV,VIND DIRECTION JOVSTICA.DIAL.SVITCA.SET VALUE.a10. a) DOOR LOAO.UP LOALMITORING
ARDEALRESP0NCES.1IME.A1O. WU( ( SMOONAUSTER.WAUDDER) OUTPUT(TARUSTER,Coo.RUDER)
Fig. 11 Schematic diagram of signal line
4. Simulation Study
4.1 Simulation System
Since the ship was not yet built, and the
control program was
being studied, the only practical way to implement the control
algorithm was to use the simulator technique.
For this
simulation, the main hydrodynamic and aerodynamic data were
obtained from model ship basin test and from wind tunnel test.
The Planner Motion
Mechanism (PMM) test,which is a kind of
forced oscillation test, was carried out about the scaled model
ship of the "Oshoro
Maru" at Akishima
Laboratories (Mitsui
Zosen) Inc. as shown in Photo.4.
The total simulation was carried out on a desktop
shiphandling simulator
(Called Harbor Master).
The simulator
communicated with a PACS substitute via a serial (RS232C)
data
link. The block diagram of this
simulator system is shown in
Fig.13. [6] SCALD FILTBRINO PILTSSINO FILTSAINO I TTTTTT CRIIMPUTI AW Cps
Fig. 12 Structure of control program I TTTTTT CO(OUTPUTI MD COMM .04,1 .A10 .I010 ()MUG WIND DIN SOT am. 7 11/11D SENSOR! DATA OVIR 000NAECTOI (0 22222 ION AM DIGITAL ,4011A(CTOI (0 ttttt 10A 801) ,,CONNECTOR 4.1 (MAINTAIIANCE) tttttAR Natritirme) 0' 102326 Tanium. OMM ()ANALOG liO ODA
CONNECTOR0 /0 ANALOGCONNECTOR COMPUTER ROOM
ST COI
VAL. MODS DIAL
&ROYCE
PLUCTU.-ATION
4.2 Ship Dynamics
A mathematical model of three principal equations, describing
longitudinal
and lateral and yaw motion, was developed for thesimulations and the control system design. The variables used
to describe the horizontal ship motions
are explained in
Fig. 14.
The principal equations are expressed as follows: m(u - vr) = XH + XE
m(v + ur) = YH + YE
Izzr =
NH + NE
The hydrodynamic forces
XH, YH,
and moment NH are complicatedfunctions of ship motion, rudder angle, and propeller thrust.
The external influences
XE, YE,
and NE are complicatedfunction
of wind, wave, and current forces. [7]
t.r
-COMMAND: BOW THRUSTER
Cpp RUDDER YAW YAW RATE SPEED. INFLUENCE: MIND CURRENT
RESPONCE: BOW THRUSTER
CDP RUDDER SIHURATION DESKTOP SIMULATOR DATA STORAGE DISC
Fig. 13 Block diagram of simulation system
Photo. 4 View of PMM test
Fig. 14 Coordinate system
CONTROL
PACS
OPERATION TERMINAL
4.3 Simulation
Results
Before implementing the PACS to an actual ship,
we tried a
control simulation. The main purpose of these
simulations were
to examine the control algorithm and
gain constants for the PACS.
Fig.15 and Fig.16 show the
course and velocity
stability
performances of the PACS in terms
of trajectory, time history
of heading,
velocity and
several actuators.
Through these
control simulations,
the control algorithm and gain constants
were reasonably
planned.
The circumstances
during the
simulations are shown in Table 2.
5. Full Scale Test
5.1 Test's Results
The PACS is
currently operatingthe "Oshoro Maru". It is
installed as a part of an
integrated system. Fig.17 shows
the result of the PACS operating the ship in
light sea condition.
These were carried out
under relative condition wind speed
averagely 10 (m/s) and
the direction of about port 10 (deg.)
bow.
This figure
shows the
ship's velocity
and course
deviation. We can see that
velocity and course keeping quality
is stabilizing.
Fig.18 shows typical result
from the sea tests in a fresh
breeze state.
This test was carried out
under the wind
condition of relative speed
until speed 10 (m/s) averagely and
the direction of port 10 (deg.) bow
was obtained. These were
wide fluctuation in direction. In this
case, the course keeping
performance was becoming wrong.
These results suggest that if
fluctuations of wind direction
are wide, the control sequence of the bow
thruster can't satisfy
the course keeping performance.
Table 2 Circumstance of simulation
HEAD(deg) SPEED(k1.8) vol. dir.(deg) vel.(las)
I dir.(deg)
0.0 0.0 10.0 350.0 0.0
0.0
350.0 0.5 10.0 340.0 1.0 I 170.0
DES 1RED INFLUENCE
VALUE
current
.wind
1130 134c1 309 540
YAW
5.2 Problems
The full scale tests carried out on PACS, indicate that a PID controller would be very difficult to implement with sufficient security over the entire range of operations for the "Oshoro Maru".
41.50 .0.13 Wm
02.W
Fig. 15 Simulation result (under wind & current)
.00 Ikts1 Vx 0 wind
\I
1011 (sic) y (YeFig. 16 Simulation result (under wind) YAW ft, .0.11 am. -0.05 Mifl. -0.1? 3.0 eve. 4.97 min. .0.03 1 Vy 4 4 Wn. 4 e I ) Vy . m, ,ti Wm 4 0 l BOW THRUSTER ...j -N ..:.:,
i'l lr m
1490/ MIL 4.01 -0.80 -10.33 P' BOW THRUSTERvim
11.11 M sad 00 7:.11 CP P . -3.03 iv* 1.32 min. 4.113 10(01 RUDDER 4.70 win. -24.62 IRO Lmr.1 5.10 720 Ideol RUDDER .3.17 -0.11 Wm -35.W 110 10931 31313 510 733 0 M WO 30 540 WO 1 tee 2.45 *0.01 140 7:0 0.12 lcu.minuu. I. 11 15tC2 0.13 x 0.211111LE/. 1.. 0 - 720 15E00 Ws, .10 .0? .W .10 03 .03
The reason for this lie in several
characteristics
asdescribed below:
* It is clear
that the pressure system of the bow thruster
of "Oshoro Maru"
is not
preliminary designed
for continuous
driving.
* Conventional PID control predominantly deals
with linear
systems within moderate disturbances.
However,
this control
approach
will not always
besatisfactory when
the operating
conditions change.
Theactuators of a bow thruster,
a Cpp and a rudder
have the
requests of the
operation with a nonlinearcharacteristics.
These create
difficulties unless special precautions are taken.
284 (deg) 254 224 2 (kts) -2. 20 (deg) 0 -20 40 -40 TAN ANGLE SPEED RUDDER 284 (deg) 254 224 2 (kts) 0 -2 20 (deg) 0 -20 YAP ANGLE SPEED BOP THRUSTER coo 40 RUDDER
77-6iztv
(deg) 0 (deg) 0 (sec) 600 (sc)Fig. 17
Full scale result( light sea )
Fig. 18
Full scale result(breeze sea)
(set course & speed 255 deg., 0.5 kts)
(set course & speed 255 deg., 1.0 kts)
6. Conclusions
The PACS for oceanographic maneuvering of the "Oshoro Maru" was designed. The performance of the PACS was investigated by
simulating studies and actual sea tests.
The PACS has the
following features:
The PACS can operate the ship and the actuators under
manual, semi-auto, and full automatic control by mean of
actuator selection buttons.The PACS can control the ship's heading and speed
automatically in the automatic mode.Joystick and dial control can adjust the ship's velocity and direction.
The PACS is operated away from the control console.
The PACS has high reliability with the adoption of a board
computer and a ROM in the processor unit.
Based on the owner's specification of the ship, as well as
the condition of the sea in the area where the vessel is to
operate, the control situation is
simulated to help in the
design and operation of the PACS.
It is concluded that the ship can be automatically operated at
wind speeds not exceeding approximate 10 (m/s) and when wind
direction is almost steady.
For wide fluctuations of wind
direction, it is necessary to modified the control sequence in consideration of the pressure system of bow thruster. Now we
are going to describe simple robust methods that can be used to
get crude estimates of automatic control system.
7. Acknowledgments
The authors are grateful to the crew of the "Oshoro Maru" for their help in implementing the PACS and their support in the
actual sea tests.
References
[1] "Data Record of Oceanographic Observations and Exploratory
Fishing", No.33, The Faculty of Fisheries, Hokkaido University, March, 1990.
H.Oda, K.Masuda and K.Karasuno, "Development of Portable Automatic System for the Fisheries Training Ship", 29th SICE
'90, Tokyo, 1990. (in Japanese)
K.Ohtsu, M.Horigome, Y.Yamanouchi, M.Hirano and H.Oda,
"Development of Energy Saving Auto-Pilot
System
through
Statistical Optimal Control", Mitsui Technical Review, Vo1.120, 1983. (in Japanese)
H.Oda, "Identification of Feed Back System through
Parametric Model", Report of UJNR-MEC
Panel, 12th Joint
Meeting, Maryland, 1983.
H.Oda, K.Masuda, K.Karasuno, K.Ohtsu
and S.Januma,
"Design of Course Keeping System at the Dead Slow Speed Applying the Multiple Auto-Regressive Model", 14th SICE System Symposium, 1988.(in Japanese)
S.Moriya and M.Hirano, "Development of a Desktop Simulator
and its Application of Simulator
Link", MARSIM & ICSM 90,
Tokyo, 1990.
K.Karasuno, K.Yoneta, S.Januma, "A New Mathematical Method
of Hydrodynamic Force and Moment Acting on Hull in Maneuvering